Good vertical transport blends math, human behavior, and the stubborn realities of architecture. If you’ve ever watched a lobby fill up at 8:55 a.m., felt the anxiety creep in as the car stops at every floor, and later sat through a fire drill that jammed the stairwells, you already understand the stakes. Getting elevators, escalators, and ramps right is one of those tasks that only gets noticed when it goes wrong. It affects leasing, code compliance, construction phasing, operational costs, and occupant satisfaction for the life of a building.
What must be moved, and when
Before choosing equipment or drawing shafts, start with movement. Buildings have pulses. Offices spike in the morning and midafternoon, residential buildings peak in the evening, hotels surge around checkout time, hospitals move around the clock but carefully, and malls pulse on weekends. Each use has characteristic traffic patterns that shape both the number of elevators and their control strategies.
In practice, planners translate these patterns into design criteria expressed as handling capacity and interval. Handling capacity is the percentage of the building population that an elevator group can move in a fixed period, commonly five minutes for office design. Interval is the average time between car departures from the main lobby. For a busy office, I typically aim for 12 to 15 percent handling capacity in five minutes with a main lobby interval in the 25 to 35 second range. Those numbers change with higher service expectations, mixed uses, or special conditions such as trading floors where short bursts of movement exceed typical office peaks.
Population estimates anchor the math. For speculative offices, 8 to 12 square meters per person is a common planning range, depending on the leasing strategy and market, which translates to roughly 80 to 125 people per 1,000 square meters. If a tenant intends dense open plans or desk sharing, or if an authority requires a conservative occupant load, the elevator count may jump by one or more cars. I have seen early elevator counts cut by half when a developer underestimated density, only to be forced into expensive redesigns later. Secure the load assumptions early, document them, and align them with the fire and life safety engineer, who may be applying different occupant loads for egress.
How building geometry shapes the solution
Elevators are not abstract objects. They live inside shafts, and shafts consume rentable area. A 30 to 40 story tower with low to mid rise expectations might use a single zone, but once you stretch beyond 25 to 30 stops, the round trip times get unwieldy and the interval suffers. Zoning the building into low, mid, and high rise groups reduces stops per car, keeps intervals tight, and lets cars travel faster over long runs. Zoning also permits localized lobbies and destination dispatch screens at transfer points, which can streamline security and wayfinding.
Ceiling-to-floor stacks matter. The structural depth, plenum space, and floor-to-floor heights dictate how much hoistway is available for gear and counterweights. Machine room-less (MRL) traction elevators have helped reclaim roof or penthouse space, yet they bring their own constraints on capacity, rise, and maintenance access. For very tall buildings, machine rooms still make sense to handle motor size, heat, and serviceability.
Core placement is the other big constraint. Centrally located cores give people short walking distances to the lifts, shorten security lines, and simplify egress. Offset cores, common in complex sites, may force long elevator lobbies or doglegged corridors that add queuing space and reduce usable floor area. In one retrofit I worked on, relocating the core was impossible, so we widened the lobby, added a gently curved wall to avoid dead ends, and introduced a third bank of escalators between office floors to absorb up-peak load. The architectural tweaks made the elevators feel faster, even though the dispatch algorithms were unchanged.
The craft of sizing cars and groups
First, right-size the car. Car capacity is typically expressed in kilograms or kilograms with a nominal person value derived from a standard body weight. A common office passenger car might be rated around 1,150 to 1,350 kilograms, 15 to 18 persons nominal, which practically means 10 to 12 adults without discomfort. Bigger cars are not always better. Large cars invite more stops as passengers are willing to board even when half full, and longer door cycles slow the group. On tight floors with high peak demand, two medium cars often beat one large car.
Door width and speed drive boarding performance. A 1.1 to 1.2 meter clear opening with center-opening doors is typical for office loads. Wider doors matter in hospitals, where beds need 1.4 meter clear openings, and in public assembly floors where flows reverse rapidly. The cycle from doors begin to open until doors close cleanly can swing by several seconds between models and settings. Those seconds pile up over dozens of stops during peak periods. Specify door hardware and controllers with the same care you give hoist motor speed.
Group size is a trade-off between average wait times and capital costs. Four to six cars per group is a sweet spot for many buildings. Smaller groups save shaft space, but they amplify the impact of a single car out of service. Larger groups can reduce average interval but may waste deliveries when traffic is light. This is where destination dispatch earns its keep by batching passengers heading to similar floors, cutting intermediate stops and tightening intervals.
Speed, acceleration, and comfort
Elevator speed is not just a number for the brochure. It must be matched to rise, stop count, and expected interval. For a 20 story office, 2.5 to 3 meters per second often works. For 40 stories, 4 to 5 m/s is common. For supertalls, you can see 8 to 10 m/s, with careful attention to pressure management. Acceleration and jerk profiles influence perceived comfort more than top speed. A car that leaps to full speed feels quick in an empty shaft but can unsettle passengers when crowded. Most manufacturers offer tunable ride profiles that can bias toward comfort or throughput. Demand realistic test rides during commissioning, preferably with a dozen people in the car during a morning peak simulation.
At high speeds, pressure differentials in the ear become noticeable. Ear popping is not just a novelty; it drives complaints. Pressure relief features, such as guide rail groove design, venting in the car, and aerodynamic hoistway features, matter. In one tower where we pushed to 7 m/s, the only item in the post-occupancy survey that drew sustained criticism was ear discomfort, not wait times. We tuned the pressure relief valves and softened the acceleration curve, and the calls stopped.
Destination dispatch and control strategy
Classic collective control lets passengers choose their floor inside the car, which leads to a scatter of stops. Destination dispatch asks for the floor at the lobby, then assigns a car that clusters riders by destination. The gains vary with building size and traffic pattern. In a 30 story office, I have seen 20 to 30 percent reductions in average travel time during up-peak and cleaner lobby queues because passengers know which car is theirs before the doors open.
The system shines when paired with security. Integrate destination screens with access control so a badge tap automatically calls a car to the correct floor range. Visitors can receive QR codes or temporary credentials tied to allowable floors. The friction reduction is real, but plan for edge cases. People leave their phones, badges demagnetize, contractors need floor-by-floor access. Place a staffed concierge or a help intercom at each destination kiosk during the first months after opening. And do not bury the kiosks behind planters or columns. If a passenger cannot find the terminal in three seconds as they enter the lobby, your queue will back up.
Use the control system to adapt to diurnal patterns. Many controllers can switch algorithms by time period: up-peak in the morning, two-way midday, down-peak in the evening, and event modes for auditorium releases. Modern systems also apply real-time learning. Those systems work best when the building population is stable week to week. In spaces with volatile loads, like convention hotels, manually tuned schedules and active monitoring during events still outperform blind learning.
Escalators, moving walks, and the vertical weave
Elevators carry people long distances efficiently, but escalators handle short vertical hops continuously. In retail podiums or transit-oriented projects, a bank of escalators can peel a significant share of elevator demand by moving able-bodied people between adjacent floors. The difference between a comfortable escalator and a source of accidents lies in sightlines and landings. Provide generous landings at top and bottom, avoid abrupt turns that force people to pivot quickly as they step off, and maintain consistent handrail speeds. If a corridor forces an immediate turn, use cues in the floor pattern or lighting to slow people before the transition.
Escalator direction matters more than many realize. In an up-peak office arrival, two ups and one down into the lobby from the parking or transit level can clear queues. Reverse that for the evening. Choose controllers that allow reversible direction and train security staff to manage the changeover quickly. On days with special events, a ten minute misalignment of direction can produce lineups that take half an hour to dissipate.
Large floor plates sometimes benefit from interior convenience stairs between adjacent floors, especially in agile office layouts where teams split across levels. These stairs absorb short-trip traffic that would otherwise load the elevators. Keep them within 10 to 15 meters of the core to feel like the natural path.
Codes, egress, and fire service
Local building codes shape vertical transport more than any other factor. The requirements cover car sizes for accessibility, fire recall behavior, firefighter’s emergency operation, and sometimes evacuation use of elevators. Coordinate early with the authority having jurisdiction. Do not assume a rule from one city applies in another with the same building type. For example, some jurisdictions allow occupant evacuation elevators that operate during an alarm under special protections: hardened lobbies with direct communication to a fire command center, redundant power feeds, smoke control, and tested control sequences. Where allowed, these elevators can trim stair widths or shorten evacuation times significantly, especially for tall towers with high occupancy.
Refuge floors or areas of rescue assistance need clear labeling and reliable two-way communication. If an emergency elevator lobby is also a security checkpoint in daily use, rehearse the sequence that clears barriers during an alarm. The worst failures I have seen in drills were not mechanical. They were operational, with locked doors or unmanned vestibules that trapped people in ambiguous spaces.
Never let elevator planning drift apart from stair egress design. Elevator lobbies double as egress paths in many layouts. Pinch points at lobby corners and interference with openable smoke doors can turn a compliant plan into a dysfunctional one. Build foam mockups or tape out lobby extents during design development. Walk the paths with the fire engineer and security team and revise where hips and shoulders actually move, not just where lines meet code diagrams.
Energy and lifecycle costs
Elevators used to be energy hogs. Modern variable voltage, variable frequency https://ads-batiment.fr/ drives, permanent magnet synchronous motors, regenerative braking, and LED lighting have changed that profile. Regenerative drives in traction elevators feed power back to the building when the counterweight assists the load, which happens a surprising portion of the time. In a midrise office with six 1,150 kilogram cars, regen can cut elevator energy use by 20 to 40 percent compared to legacy gearless machines, depending on traffic patterns and duty cycles.
Standby modes matter. Car fans and lighting should drop to low consumption when idle, doors should not cycle unnecessarily, and the controller should park cars efficiently during off-hours. I still specify separate submetering for elevator and escalator power. Over a year, metering reveals real duty cycles and supports tuning. Without metering, arguments about savings are theoretical.
Lifecycle thinking must include maintenance access and part availability. An exotic imported machine with proprietary tools may look great on day one, but if the local market has two technicians trained to service it, your downtime risk climbs. Standardize components across groups where possible, stock critical spare parts on site for large buildings, and include maintenance response times as a contractual performance metric. If a car goes out during peak, specify an escalation protocol that triggers dispatch adjustments or security support for queue management.
Construction phasing and modernization
On new projects, staging the installation is an act of diplomacy among trades. Shafts become vertical highways for materials and workers long before the cars are finished. Plan for temporary lifts or early handover of one shaft with a construction-rated car. If the general contractor treats the car as a freight lift for drywall pallets without proper protection, you will inherit a maintenance headache. Protect finished cars with hoardings and limit loads according to the manufacturer’s instruction. Schedule a mid-construction inspection with the supplier to catch damage early.
Modernization in occupied buildings is harder than new work. Taking two cars out of a six-car group for six months can wreck service levels and tenant patience. I favor a rolling modernization: one car at a time, starting with the car that will become the template for the rest. Commission it fully, test the dispatch, tune the door hardware, then apply lessons to the next. If your building has low and high rise groups, stagger shutdowns so no one population suffers. Communicate clearly. Post expected wait times in the lobby during peak, explain progress weekly, and give tenants a number to call that reaches a person with authority to act. People tolerate inconvenience when they believe someone is on the other end who can fix it.
The human factor
Elevators move people, and people do unpredictable things. They hold doors, they crowd the threshold, they press buttons twice, they take strollers and carts they were not supposed to bring. Good planning anticipates behavior and designs for it. Visible countdown indicators for doors reduce futile dashes. Clear signage that shows which floors a group serves prevents dead-end rides to transfer points. In mixed-use buildings, separate residential and office lobbies reduce conflict and security incidents. Where populations mix in a shared lobby, distribute destination kiosks logically: one at every major entry stream, aligned with natural walking vectors.
Security screening throws a wrench into smooth flows. Bag checks at elevators can double wait times if not staffed appropriately. If the building requires screening during certain hours, design queue space with more length than you think you need. A rule of thumb I use is 0.8 to 1 square meter per person in queue for brief checks and up to 1.2 square meters where lines last longer. Test the queue during a soft opening, not on day one with full occupancy.
Special occupancies: hospitals, hotels, and labs
Hospitals are a different world. Separate service and patient flows, oversized beds, infection control barriers, and 24/7 duty cycles call for dedicated car types and robust redundancy. A general acute care hospital may have three to four passenger groups plus dedicated service cars sized for beds and equipment. Door protection with extended light curtains, drip-resistant features, and antimicrobial finishes help. Align elevator lobbies with nurse stations and critical departments to minimize lateral travel. And never let construction phasing close off the only bed elevator serving a surgical suite without a tested workaround.
Hotels blend public and back-of-house. Keep guest elevators visible from the front desk to ease wayfinding. Provide a discreet service group that reaches every floor with pantry and housekeeping spaces aligned to reduce travel distance. If the building includes ballrooms, expect heavy movement at predictable times. Consider event modes that allocate more cars to high-floor guests during checkout while preserving some capacity for banquet crews moving equipment.
Labs and research spaces sit between office and hospital. They often require service cars for gas cylinders and equipment that cannot travel in passenger cars, and they may need zoned access with strict security. Plan staging rooms at elevator lobbies for safe handling of cylinders and waste. Where hazardous materials are present, consult with the safety officer early about elevator shaft pressurization, smoke control, and spill response.
Digital twins, simulation, and testing
Traffic simulation tools are useful, but they are only as good as the inputs. Feed them with credible occupant loads and arrival profiles. Avoid the temptation to chase one magic number like average waiting time; look at distributions. A system with a low average can still have an unacceptably high tail where the last 5 percent of passengers wait too long. Include wheelchair users in the simulation, especially where the building has heavy public traffic. A destination dispatch algorithm that packs cars to 90 percent capacity may inadvertently discourage boarding by users who need extra time, increasing their delay.
When the hardware is installed, stage live tests with real people. Advertise a “peak test” day with gift cards and snacks to encourage participation. Measure both wait times and perceived satisfaction. I have watched a group pass its numeric targets only to earn poor ratings because the lobby acoustics amplified the door chimes into an irritating whine. An acoustic fix and a chime frequency adjustment improved satisfaction overnight without touching the dispatch.
Ownership, operations, and the long view
Elevators are long-lived assets. A good installation runs 25 to 30 years before major modernization, with proper maintenance. The largest operating costs are service contracts, parts, and energy. Create a maintenance regime that is specific, not generic. Define preventive maintenance intervals tied to cycles, not calendar days alone. Track calls and fault codes with trend analysis, not just work order closures. If a door zone sensor triggers twice a week on the same floor, inspect the sill and threshold, not just reset the controller.
Train building staff. A facilities team that understands basic fault codes and reset procedures can prevent half-day outages. Provide a quick reference that lists which faults require a vendor and which can be safely cleared by trained in-house staff. Encourage a culture of reporting among occupants. A note about a door closing too fast or a rough stop can signal a drift in settings before it turns into a shutdown.
Finally, keep an eye on technology changes that truly add value without locking you into proprietary ecosystems. Open protocol monitoring platforms can aggregate data across manufacturers and buildings. Smartphone call apps can reduce touchpoints and improve accessibility for some users, but they should not replace physical call devices. Touchless does not mean screen-only. Maintain tactile buttons, auditory cues, and consistent signage for people with low vision or limited dexterity.
A compact planning checklist
- Clarify population loads by use and time of day, document assumptions, and align with egress design. Map building geometry, core location, and zoning to minimize stops and travel distance. Select car sizes and speeds tuned to rise and comfort, not just peak marketing numbers. Integrate destination dispatch with security, and place kiosks where people intuitively look. Plan maintenance, modernization phasing, and spare parts strategy from day one.
What success looks like
A well-planned system feels boring in the best way. The lobby flows without drama during the morning rush. People step into cars that are busy but not crushed. A service car handles a pallet delivery without blocking passengers. A destination screen that failed last night has already been swapped by midday, and the backup call panels carry the load without confusion. On drill day, the occupant evacuation elevators behave exactly as the fire marshal expects, and the stairwells never jam. In a year, the energy meter on the elevator panel shows a consumption curve that matches the original model, which let the owner adjust lighting and fan standby settings to trim another few percent.
You get to that point by treating vertical transport as infrastructure with human behavior at its core. Start with honest movement data. Let the building’s geometry inform the strategy. Respect the codes, but also the people who will ride the cars every day. Test, tune, and listen. Elevators are the arteries of a building, and like any circulatory system, they work best when you design for the rushes, the lulls, and the surprises.